PDGM Data Shows Reality-Versus-Expectations Mismatch for Case-Mix Components

The Centers for Medicare & Medicaid Services (CMS) was a little bit off with some of its predictions related to the Patient-Driven Groupings Model (PDGM). 

At least, that’s true when it comes to the first four months of 2020, according to Strategic Healthcare Programs (SHP) data unveiled on a recent BlackTree Healthcare Consulting webinar.

The webinar covered data and trends related to the impact of PDGM and COVID-19 on home health agencies. As part of the presentation, experts compared SHP data to CMS projections on the various components of PDGM.


CMS detailed those projections in its 2020 final rule, using 2018 claims data paid through July 31, 2019, to inform its calculations. When put up against SHP data from January through April, it appears some of CMS’s predictions have yet to align with reality.

Take comorbidity adjustment, for example. It’s one of the five levers used to determine which of the 432 case-mix groups a case falls into under PDGM — and in turn, the reimbursement level that goes with it.

Under the new framework, there are three different comorbidity buckets: no adjustment, a “low” adjustment” or a “high” adjustment.


According to its models, CMS expected 56.4% of home health cases would have no adjustment; 35.5% would have a low adjustment; and only 8.1% would have a high adjustment.

In reality, 14.1% of cases had a high adjustment in the first four months of the year, according to SHP data. Those high adjustment cases are associated with a combination of secondary diagnoses that are tied to higher resource use when reported together — thus yield higher reimbursement.

“Some of that might be we’re doing a better job at secondary diagnosis coding, we’re paying more attention to that [or] we’re seeing more codes on the claim versus maybe what was just put on the OASIS,” Chris Attaya, vice president of product strategy at Strategic Healthcare Programs (SHP), said during the webinar.

Meanwhile, the expectations-versus-reality mismatch in functional impairment lever came as the biggest shock, according to BlackTree Managing Principal Nick Seabrook.

“The one area that surprised me the most is just the disparity in terms of the functional impairment scores,” Seabrook said during the webinar.

Like comorbidity adjustment, PDGM’s functional impairment lever is broken into three categories — low, medium and high — each of which comes with varying degrees of reimbursement. A higher impairment generally means higher reimbursement.

While CMS anticipated a fairly even split between all three functional categories, nearly 44% of cases have fallen into the high category. That’s over 12% more than what CMS expected, according to the data.

Source and timing trends are also slightly different from CMS’s expectations. Generally, the SHP data suggests that the industry is seeing far more “institutional early” cases and far fewer “community late” than CMS expected.

Specifically, CMS projected 61.4% of cases would be community late, while SHP found that only 54.4% were. Additionally, CMS estimated that 18.5% of cases would be institutional early, compared to 27.4% in reality.

Attaya wasn’t surprised by that mismatch, however.

“We’re seeing a higher [percentage] of early because … we’re still new with PDGM,” Attaya said. “Community late — we haven’t quite caught up yet, but I think we’ll see numbers continue to get closer to CMS [projections] as we go forward through the year when we have more opportunities for recerts that would continue to grow that ‘community late’ period.”

Other subcategories under the source and timing lever include “institutional late” and “community early,” which were on par with CMS’s expectations.

While the data provides a lens into the first four months of PDGM, it only tells part of the story. Home health providers were grappling with the COVID-19 emergency for the last several weeks of that four month period.

“As we kind of look at some of the trends, you’ve got to remember that [COVID-19] is included in there,” Attaya said. “That’s going to have some impact as you look at your overall numbers.”

Companies featured in this article: